Politics Close to Home: The Impact of Meso-level Institutions on Women in Politics

Candice D. Ortbals*, Meg Rinckery, and Celeste Montoyaz

*Pepperdine University; [email protected] y Purdue University Calumet; [email protected] Downloaded from zUniversity of Colorado; [email protected]

Scholars recognize a worldwide increase in decentralization as well as the prevalence of multilevel

governance in Europe. This article examines the advantages and disadvantages that meso-level http://publius.oxfordjournals.org/ institutions present for women’s political representation in three European Union member-states that are decentralized, unitary states. Using the framework of the triangle of women’s empower- ment, we ask whether women are represented in meso-level legislatures, women’s policy agencies, and women’s movements in , Spain, and Poland. We find that gains in meso-level legislatures are slow, but meso-level women’s policy agencies and movements provide important access for women to politics. Like scholars studying women and , we conclude that decentralized institutions in unitary states offer both opportunities for and impediments to fem- inist policy and activism. at University of Colorado on March 13, 2012

Current discourses about decentralization promise political participation, repre- sentation, and policy for women citizens. For example, in Spain, a sub-state government institute charged with improving lauded itself as an institution ‘‘with a new philosophy ...where women as individual subjects and as a collective achieve full participation’’ (Delegacio´n de la Mujer 2003, 19). Nevertheless, evidence from other countries suggests that subnational governments are not providing women greater political participation and rights. South Korean feminists, facing low representation in national institutions, emphasized success in local elections, hoping success there would ‘‘spillover’’ into the national sphere. Disappointed with local politics, they began targeting the national level again (Chin 2004). Moreover, in 1999, a regional autonomy law gave wide power over social and economic policy to regional governments in Indonesia, and conservative ones responded by imposing strict dress codes for women (Siahaan 2003). Decentralization and federalization became prevalent characteristics of the modern state in the latter decades of the twentieth century (Elazar 1991;

Publius:TheJournal of Federalism volume 42 number1,pp.78^107 doi:10.1093/publius/pjr029 AdvanceAccess publication August 5, 2011 ß TheAuthor 2011.Published by Oxford University Press on behalf of CSFAssociates: Publius, Inc. All rights reserved. For permissions, please email: [email protected] Impact of Meso-level Institutions on Women in Politics 79

Watts 1999; Falleti 2005). Sub-state institutions historically have allowed strong regional identities to assert themselves, and more recently they have elicited hopes for increasing citizen participation (Cain, Dalton, and Scarrow 2003), reducing the obligations of the central state (see Banaszak, Rucht, and Beckwith 2003), or winning partisan battles for electoral dominance (O’Neill 2003). As women in most societies disproportionately work in paid and unpaid care professions (Folbre 2001), political outlets situated geographically closer to work may facilitate women’s representation. This is important because women lack representation in national political institutions; as of 2009, they constituted 18.5 percent of national Downloaded from legislators worldwide, thus additional arenas for representation pose great advantages for them (Interparliamentary Union 2009). Alternatively, federalism scholars find that meso-level institutions, though offering opportunities to feminists, impede their success by requiring them to contend with multiple http://publius.oxfordjournals.org/ governance sites (Chappell 2002). Therefore, scholars ‘‘need to consider the gender-specific implications of decentralization,’’ both advantages and disadvan- tages (Rai 2003, 35). Scholarship about gender politics and subnational institutions in historic federations is instructive (Gray 2006; Chappell 2002; Vickers 2010), yet, according to Gray, it ‘‘is missing important [comparative] evidence’’ by failing to examine decentralized, unitary states—some of which may experience policy decentralization as far-reaching as federalized states (2006, 38). As such, there is a need for cross-national, cross-institutional, comparative research in recently decentralized at University of Colorado on March 13, 2012 countries examining whether women are present in meso-level political institutions and whether institutions further and/or impede feminist policies and activism. Thus, we ask how do meso-level institutions influence women’s representation in politics in recently decentralized countries? This comparative case analysis examines meso-level units between municipal and national administrative levels, in the three most-similar European Union countries of Italy, Spain, and Poland, all of which experienced decentralization from the 1970s to 2000s and, currently, multilevel governance in which European Union, national, provincial/regional, and local governments influence public policy. We evaluate whether bringing ‘‘politics close to home’’ improves women’s representation through Lycklama a Niejholt et al.’s (1998) concept of the ‘‘triangle of empowerment,’’ that includes (i) women legislators (ii) women’s policy agencies, and (iii) women’s movement organizations, whose collaboration is associated with the development of feminist policies. Relative to each node of the triangle, we report women’s representation by documenting whether (i) meso-level legislatures yield 30 percent women (estimated by the United Nations to ‘‘fairly represent’’ women, see Lovenduski 2001), with more women serving in meso-level than national legislatures (ii) meso-level women’s policy agencies engage in feminist collaborations and further feminist policies expected of European member-states, 80 C. D. Ortbals et al. including employment, work-family balance, and anti-gender violence, and (iii) women’s organizations establish themselves and work outside national capitals, in the countries’ peripheries. Our analysis shows that actors in the triangle of empowerment do not always act in concert to develop feminist policies and, contrary to popular belief, women’s descriptive representation is sometimes lower in meso-level than national legislatures, not reaching 30 percent women representatives unless specific quotas for women are passed to ensure women’s representation. Under decentralization, however, women’s representation increases through meso-level women’s policy Downloaded from agencies that tend to address feminist policy and women’s movements that have become more numerous in the peripheries since decentralization. Similar to studies of women’s representation in federal states, we conclude that meso-level institutions hold significant potential but do not constitute a panacea. Moreover, http://publius.oxfordjournals.org/ we find notable variation among our most-similar decentralized states just as scholars have found variations between most-similar federal states (Chappell 2002). Thus, in the conclusion, we begin a conversation about the benefits posed by decentralized states, what variables facilitate said benefits, and the variations among seemingly similar states that beg for more comparative studies of ‘‘politics close to home.’’ The article is organized in four sections: terminology and case selection, theory review, comparative case analysis, and conclusions. at University of Colorado on March 13, 2012 Terminology and Case Selection It is important to clarify definitions of meso-level institutions, decentralization, federalism, , and the triangle of empowerment. To not conflate decen- tralization and federalization, we use the term ‘‘meso-level institutions’’ as shorthand for institutions resulting from sub-state dispersal of power. In broad terms, decentralization is the ‘‘assignment of fiscal, political, and administrative responsibilities to lower levels of government’’ (Litvack, Ahmad, and Bird 1998, 4; Litvack and Seddon 2002; Rodden 2004). Federalism is ‘‘an arrangement of shared-rule through common institutions and regional self-rule for the govern- ments of the constituent units’’ (Watts 1999, 7). Table 1 shows that all three countries experienced decentralization reforms between the 1970s and 2000s. Fiscal, political, and policy/administrative decen- tralization has occurred in Italy, Spain, and Poland. Each country demonstrates moderate levels of fiscal decentralization. As of 2005, these meso-level governments are responsible for collecting around 15–30 percent of revenues, showing that central governments remain influential over regions. Our cases fall near the 30–50 percent for meso-level expenditure as a percentage of total expenditure, with Italy, Spain, and Poland respectively with 31, 54, and 31 percent, compared to Denmark with 65 percent, thus meaning that meso-units in our study manage Table 1 Case selection table of European, Catholic countries with decentralization to meso-level units Politics in Women on Institutions Meso-level of Impact

Italy Spain Poland

Number of Meso-level Units 20 17 16 Timing of decentralization 1970–1990 1980–1990 1998–2008 Political decentralization Directly elected regional legislators ˇˇ ˇ Directly elected regional governors ˇˇ Regions created by constitution ˇ Regions protected by constitution ˇˇ Fiscal decentralization Sub-national expenditure as a percentage of total 31% (2007) 54% (2007) 31% (2007) expenditure Sub-national revenues 15% (2005) 31% (2005) 13% (2005) Administrative Decentralization Regional Women’s Policy Agencies ˇˇ ˇ Regional-Focused MLG on Economic ˇˇ ˇ Development GDP (Per capita PPP) 30,631 (2008) 30,589 (2008) 17,537 (2008) Religiosity (0–3Pew Scale) 0.8 (2007) 0.7 (2007) 1.0 (2007) Gender Empowerment Rank 209 (1–105) 21 (2007) 11 (2007) 41 (2007) National electoral system Closed-list proportional Closed-list proportional Open-list proportional representation representation representation Regional electoral system Open-list proportional Closed-list proportional Open-list proportional representation representation representation National gender quotas Voluntary political party Legally required candidate Voluntary political party quotas quotas quotas (until 2011)

Notes: For fiscal decentralization measures, see Northern Ireland Assembly 2008. For religiosity measures, see Wolfe (2008). 81

Downloaded from from Downloaded http://publius.oxfordjournals.org/ at University of Colorado on March 13, 2012 13, March on Colorado of University at 82 C. D. Ortbals et al. considerable resources to spend on policy (Northern Ireland Assembly 2008). Political decentralization is stronger in Italy and Spain than in Poland. Italy, Spain, and Poland all share directly elected meso-legislators, but meso-units are constitutionally protected and sub-state executives are directly elected in Italy and Spain and not in Poland. All three countries have experienced decentralization of policy areas pertinent to women’s rights and empowerment, including health, education, and employment/labor policies. However, Polish municipalities and counties are responsible for education policy and many social services, and they receive more government funds than regions, which are mainly influential in health Downloaded from policy (Swianiewicz 2002). Spain, however, has experienced such considerable policy decentralization that it is more decentralized than many federal states.1 Decentralization of the aforementioned policy areas impacting women’s rights may be found in federal states as well; yet, unitary states, especially Poland without http://publius.oxfordjournals.org/ constitutional guarantees, may attempt to repossess them, which is not feasible in federations (see Breton 2000). As decentralized structures vary, so too do definitions of feminism. We consider feminism a movement against with the goal of dismantling male dominance over women in public and private spheres so that women can practice self-determination in their social, political, and economic lives (Mazur 2002; Beckwith 2000; Gelb and Palley 1982). Women’s policy agencies (WPAs), present in 127 countries worldwide (Rai 2003), connote ‘‘any structure established by government with its main purpose being the betterment of women’s social status’’ at University of Colorado on March 13, 2012 (Stetson and Mazur 1995, 3). Through persuasion of ministers and legislators, women’s policy agencies develop equality policy plans and policy. WPAs, equality plans, and gender equality policy may be feminist when they seek significant changes in gender relations (such as the elimination of sex discrimination or the empowerment of women in areas of employment and education) and otherwise engage the demands of feminists (Mazur and McBride 2008). The ‘‘triangle of empowerment’’ (Lycklama a Niejholt et al. 1998) stresses the role of women in legislatures, bureaucracies, and social movements in the development and implementation of gender equality policy. The triangle ‘‘articulates women’s demands, translates them into policy issues and struggles to widen support for their agenda. The dynamism created between these actors accounts for the relative effectiveness with which women’s interests can be defended’’ (ibid., 3–4). Triangle actors form ‘‘strategic partnerships’’ (Halsaa 1998, 183) and interact in ‘‘a friendly and open’’ way (183), so as to ‘‘lend each other expertise in policy matters’’ and develop policy together (Ortbals 2008, 98). Because scholars have not established sufficient conditions for the triangle’s dynamism, we evaluate the representation of women through the presence or absence of women in the nodes of the triangle, in legislatures, women’s policy agencies, and women’s movement organizations. Where data are available, we report whether these Impact of Meso-level Institutions on Women in Politics 83 meso-level institutions are feminist in orientation and whether they are working in partnership to achieve feminist policy. For instance, we report whether meso-level WPAs exists, if they have issued equality plans and facilitated gender equality policy, and if said plans and policies demonstrate, rhetorically and in action, commitment to feminism. Italy, Spain, and Poland do not share identical economies and sub-state arrangements, yet as we report in table 1, they are most-similar on account of their sub-state reforms, Catholic heritage, gender regimes, populations, number of sub-state units, and proportional representation electoral systems.2 They have all Downloaded from decentralized recently—Italy from the 1970s to 1990s, Spain beginning in the 1980s, and Poland in the late 1990s and early 2000s. Moreover, Italy, Spain, and Poland have higher than average levels of religiosity among European

countries. Italy, Spain, and Poland rank at 0.8, 0.7, and 1.0, respectively on the http://publius.oxfordjournals.org/ Pew Research Center scale of religiosity, compared to France at 0.3 (0 is low religiosity and 3 is highest). Our countries also share social welfare policies that emphasize women’s motherhood (Moreno Mı´nguez 2006; Fodor et al. 2002). The sizes of their population are similar (Italy—58 million; Spain—40 million; Poland—38 million) as is the number of their meso-level units (Italy 20; Spain 17; Poland 16). All three countries utilize proportional representation electoral systems that are associated with increased legislative representation 3 for women. at University of Colorado on March 13, 2012 Finally, Italy, Spain, and Poland are all member states of the European Union, which has developed a system of multilevel governance, in which ‘‘supranational, national, regional and local governments are enmeshed in territorially overarching policy networks’’ (Marks 1993, 402). Our countries became members of the European Union at different times (Italy—founding member; Spain—1986; Poland—2004), yet they are all influenced by the EU’s gender equality initiatives. The EU hastened policy change in countries that had been less supportive of gender equality such as Spain, Italy, and Poland, and it motivates the development of WPAs. The EU’s binding legislation on gender equality, developed from the 1970s to the 2000s, calls for comprehensive reform of equal employment practices, requiring member and candidate states to address various forms of gender discrimination (i.e., discrimination in hiring, pay gaps, , etc.). Furthermore, the EU has begun to expand its gender equality agenda beyond the employment realm to include newer initiatives, although nonbinding, that work to balance family responsibilities between women and men, combat , and increase women’s political representation (Lombardo and Meier 2008; Montoya 2009). As such, meso-units in our cases share a comparable context for us to examine the feminist policy areas of equal employment, work–family balance, and violence against women. 84 C. D. Ortbals et al.

Theory Review: Meso-level Advantages and Disadvantages Significant research is emerging on women’s access to subnational politics in federal and decentralized, unitary states. Below we review studies on women in the triangle of empowerment, in meso-level legislatures, women’s policy agencies, and women’s movements, noting the advantages and disadvantages they afford.

Advantages of Meso-level Institutions for Women Citizens

Meso-legislatures present opportunities such as higher numbers of women in office, Downloaded from namely greater women’s descriptive representation, which ultimately may lead to the legislation of feminist policy, or feminist substantive representation. Two factors that increase women’s descriptive representation in legislatures are proportional electoral laws and gender candidate quotas that specify minimum percentages of women on party lists or in legislative seats. Proportional list electoral systems are http://publius.oxfordjournals.org/ associated worldwide with twice as high a percentage of women in office than majoritarian systems because parties in proportional systems have an incentive to balance tickets with women and minorities to represent different constituencies (Rule and Zimmerman 1994; Vengroff, Nyiri, and Fugiero 2003). Likewise, quotas, mandated through election laws or voluntarily implemented by political parties, increase the percentage of women in national and meso-level legislatures (Krook 2009). Though theory does not purport a greater likelihood of proportional representation and quotas in meso-level units, some units with the power to revise at University of Colorado on March 13, 2012 electoral rules may adopt quotas on their own and women may more easily gain descriptive representation due to meso-level legislatures’ reputation as lower status and stakes posts that marginalized candidates can win and their geographic proximity, particularly in geographically small states, to women’s familial responsibilities (Darcy, Welch, and Clark 2003). Decentralization also likely increases legislators’ attentiveness to marginalized groups, including women, because decision-making is pushed to a unit where constituents are fewer and geographically closer, thus making it cheaper and easier for officials to ascertain and respond to constituent preferences. Meso-level units may also serve as laboratories for innovations in gender equality policies, and thus substantive representation (see Rom, Peterson, and Scheve 1998). Women’s policy agencies potentially increase women’s substantive representa- tion as they liaise across government ministries to promote gender policy and network with women’s organizations (Weldon 2002; Stetson and Mazur 1995). Subnational WPAs emerged in the late 1970s and exist in diverse countries, such as the United States, Australia, Spain, Italy, Germany Poland, Brazil, Chile, India, and Pakistan (see Rincker and Ortbals 2009). Meso-level WPAs may partner with feminists organizations nearby (Halsaa 1998), facilitate policies that honor local identities such as job training and education that relate to regional economies and Impact of Meso-level Institutions on Women in Politics 85 languages (Ortbals 2008), or promote innovative policies that ‘‘trickle up’’ to the national WPA (Macaulay 2006). Meso-level institutions also likely impact women’s movements because they typically encourage social groups, which were formerly marginalized, to gain a repertoire of participation skills and to feel more efficacious (Putnam 1993; Brady, Lehman Schlozman, and Verba 1999). If these relationships hold true for women, more women’s organizations may emerge upon decentralization and impact new institutions. Feminists may accomplish movement goals as they participate in synergistic policy deliberations (Putnam 1993; Ostrom 1996). Downloaded from Many of these advantages prove true in regional or single-case studies of decentralization. Celis and Woodward (2003) show that in half of European countries with regional legislatures, more women are elected to regional legislatures than their respective national legislatures. Vengroff, Nyiri, and Fugiero (2003) show http://publius.oxfordjournals.org/ that 63 percent of advanced democracies have a greater percentage of regional than national women parliamentarians, with women regional parliamentarians outnum- bering national ones in plurality/majority systems as well as proportional electoral systems. Furthermore, Outshoorn and Kantola find that the growing number of meso-level WPAs in advanced democracies in the 1990s and 2000s is associated with the adoption of gender equality policies in many kinds of sub-states, for instance, in federalized Belgium, weakly federalized Austria, and regionalized France (2007). Brazilian sub-state WPAs have served as ‘‘an effective space for at University of Colorado on March 13, 2012 debate’’ for the women’s movement (Alvarez 1990), and, during the 1980s, Norwegian feminists led many local institutions and pursued women’s issues there (Halsaa 1998). Finally, much scholarship suggests that leftist parties—and presumably meso-level institutions led by them—support women’s issues, empower WPAs, and ally with feminist movements (Stetson and Mazur 1995; Waylen 2007; Rincker and Ortbals 2009). Federalism in Australia muted the effect of ‘‘the election of conservative governments at one level’’ as gender advances were furthered at the other level (Sawyer 2002, 175). Similarly, in the early 1980s, British local governments were avenues for feminist activism during Tory governance (Rowbotham 1996).

Disadvantages of Meso-level Institutions for Women Citizens Meso-level legislatures are not immune from pitfalls apparent in national legis- latures. Even if more women enter meso-level legislatures, they may not constitute a ‘‘critical mass’’ required to ‘‘fairly represent’’ women (estimated by the United Nations at 30 percent of a legislative body—Lovenduski 2001, 2) or be ‘‘critical actors’’—feminist men or women who mobilize for feminist policies (Childs and Krook 2006). Moreover, masculine leadership and organizational norms may dominate legislatures. 86 C. D. Ortbals et al.

Meso-level WPAs may fail to represent women because civil servants in the subnational sphere are often less trained than national ones (Prud’homme 1995), they may be unaware of international good gender practices, and they may be unresponsive to feminists. Additionally, the sub-state may become a laboratory in which organizations fighting against women’s rights succeed (Haussman 2005). Given that gender equality policy is not the foremost policy preference in democratizing countries (Waylen 1994), it is also a public good vulnerable to budget cutbacks. Meso-level WPAs may also bear the brunt of neo-liberal ‘‘downloading,’’ namely reducing central government expenditures on social Downloaded from welfare policies, thus pushing responsibilities for gender policies to sub-state insti- tutions lacking sufficient resources (see Banaszak, Rucht, and Beckwith 2003). Like national WPAs (Waylen 2007), meso-level WPAs will be weak if they are not created by law and/or if alternating party governance abruptly changes their http://publius.oxfordjournals.org/ objectives. Decentralization also presents significant challenges for women’s organizations. Fighting patriarchy at multiple governance sites could strain organizational resources and fragment movements (see Chappell 2002). Moreover, feminists may prefer centralized gender policymaking that yields consistent policy across a country (see Sawyer 2002). Empirical research demonstrates some of these disadvantages. Polish women in regional legislatures face a choice of downplaying feminist attitudes to fit in with male leadership or being sidelined as feminist extremists (Rincker 2009) and, in at University of Colorado on March 13, 2012 Finland, a ‘‘gender equality illusion’’ incorrectly associated the high number of women in elected politics with the accomplishment of gender policy (Holli and Kantola 2007, 97). Parry finds that state WPAs in the United States lack resources and commitment by state officials (2005), and meso-level WPAs, even those under leftist governments, are constrained in assisting women’s movements ‘‘in the neoliberal context of state restructuring’’ (Sauer 2003), which Teghtsoonian and Chappell confirm has occurred in Canada and Australia (2008). Finally, meso-level institutions disappoint activists. United States abortion rights activists faired better in the nationalized policy context of the 1970s and 1980s than the ‘‘new federalist period (since 1989),’’ in which pro-life activists prospered (Haussman 2005, 66). Likewise, Australian feminists in decades past preferred national politics that provided ‘‘uniform social provisions and gender equality’’ countrywide rather than variable state-level politics (Sawyer 2002).

Comparative Case Analysis For Italy, Spain, and Poland, we evaluate whether women are present in the nodes of the triangle of empowerment. We report (i) the percent of women legislators in meso-level legislatures, noting whether/when women legislators have reached Impact of Meso-level Institutions on Women in Politics 87

‘‘fair representation’’ at 30 percent; (ii) the establishment of national and meso-level women’s policy agencies, detailing whether/when WPAs have produced gender equality plans and supported, rhetorically and in action, feminist policies; and (iii) the number of women’s organizations, especially those fighting gender violence, to show whether women’s organizations are present in all meso-level units. We examine gender violence organizations because of available data and because violence is a critical and comparable feminist issue, for women facing gender violence need rapid responses that are ‘‘close to home’’ and the EU encourages member-states to fight gender violence.4 Downloaded from WPA websites, secondary sources, and newspapers provide data to gauge WPA existence, presence of equality plans, and WPA degree of feminism. We gauged WPA degree of feminism (High, Moderate, or Low) by three indicators (i) rhetor- ical support of three feminist policies (equal employment, work- family balance, http://publius.oxfordjournals.org/ and violence against women), (ii) enacting three feminist policies (equal employment, work- family balance, and violence against women, all of which are central to EU legislation or its aspirations for gender progress), and (iii) collab- orating with women’s organizations. We rank WPA Rhetorical support for feminism (Indicator 1) High when WPAs publicly discuss all three policies through websites and publications, Moderate when they address two, and Low when they address one. Similarly, we rank WPA levels of feminist policy implementation (Indicator 2), High when they enact policies in all three areas, Moderate when they enact policies in two areas, and Low when they enact policy in one area. Finally, we rank WPA at University of Colorado on March 13, 2012 collaboration with women’s organizations (Indicator 3) High when the WPAs both (i) incorporate women’s organizations into their leadership structures and (ii) part- ner with them on programs; Moderate when WPAs engage in either leadership incorporation or joint programs, and Low when WPAs only informally communicate with women’s organizations. The Women Against Violence Europe network (WAVE) provides the number of women’s anti-violence organizations in Italian, Spanish, and Polish regions. For Italy, US State Department Human Rights Reports provide earlier data to show growth in antiviolence organizations. Additional sources determine geographic dispersion of Spanish and Polish organizations because similar reports are unavailable. Given the slightly more extensive literature purporting the abovementioned advantages of meso-level institutions, we expect the expansion of institutions ‘‘close to home’’ to positively impact women’s representation. Therefore, we expect to find (i) more women serving in meso-level than national legislatures, with meso-level legislatures reaching fair representation at 30 percent women; (ii) the presence of meso-level WPAs with moderate to high degrees of , policy, and collaboration; and (iii) women’s organizations dispersed across countries, active not solely in national capitals and providing local feminist responses to gender violence. 88 C. D. Ortbals et al.

Italy Despite a vibrant that emerged in the 1970s, 50 years of government domination by the Christian Democrats (DC) limited women’s representation and helped preserve the male-dominated nature of Italian politics. After realignment of the national party system in the 1990s, the left broke the DC’s political hold (from 1996 to 2001) and pursued women’s rights reforms in the late 1990s and early 2000s; yet reforms have been characterized by uneven implementation and inconsistent gains due to the instability and inefficiency of Downloaded from Italian governance. Fluctuation in party politics also influences decentralization. Support for decentralization began with the inclusion of provisions for elected regional government in the Italian Constitution of 1948; however, opposition from DC forestalled its implementation over twenty years. In 1970, under growing pressure from the Communist Party, the first regional elections were held. With http://publius.oxfordjournals.org/ the passage of the ‘‘616 decrees’’ in 1977, administrative powers in the areas of agriculture, health, and social service policy were transferred to the regions (Putnam 1993). Continued reform was on the agenda in 1980, but no new provisions were adopted until the late 1990s and early 2000s (Guadagnini 2007, 175). Reform included the direct election of regional presidents and more regional involvement in a number of policy areas, including equal opportunities (Guadagnini 2007). The center-left Olive Tree Coalition government provided a more hospitable environment to gender equality and added a 2001 constitutional at University of Colorado on March 13, 2012 law on regions stating that regional laws should promote equal opportunities in elective positions, presumably through regional quotas (Guagdanini 2007). Nevertheless, women’s presence in Italian legislatures is incredibly limited both nationally and regionally, hence the oft-heard claim that Italy hosts ‘‘a highly male-dominated decision-making arena’’ (Guadagnini and Dona` 2007, 165). Italian national politics has been associated with low levels of women in office, generally under 10 percent during the 1980s and 1990s. Representation increased when quotas were used for the first time in the 1994 national election and the municipal, provincial and regional elections of 1995; however, the Constitutional Court, which has been a rather conservative force in regards to gender equality, ruled these quotas unconstitutional in 1995. This precedent held until constitutional reforms were made in the early 2000s; the aforementioned 2001 regional reform specified equal access for men and women in regional office and a 2003 constitutional reform opened the door to the enactment of quotas by stating that the Republic must adopt measures to promote equal chances for men and women in elective office. In the 2006 election, the victorious Union leftist coalition (previously ) helped increase the number of women in the Chamber of Deputies to 17.3 percent when party members voluntarily enacted quotas ranging from 40 to 50 percent. In 2008, this number rose to 21.3 percent; while this is an Impact of Meso-level Institutions on Women in Politics 89 improvement, it is still below the UN’s conception of fair representation. At the regional level, six regions implemented election law quotas in 2005 elections: Abruzzo, Calabria, Lazio, Puglia, , and Valle d’Aosta. The impact of quotas is evident in comparing the average percentage of women councilors in regional parliaments over time: in the 1995 election with quotas it was 12.9 percent, in the quota-absent 2000 election it dropped to 8.8 percent, and in the 2005 election with quotas in some regions, the percent increased to 12.8, still below the 1995 level and fair representation at 30 percent (Bonomi et al. 2006). Despite these modest gains, regional representation lags behind national representation and Downloaded from only several regions have comparable or stronger representation (see table 2 and Supplementary Appendix). Unlike legislatures, WPAs offer Italian women more reasonable representation.

The first two WPAs in Italy were the National Committee for Equal Opportunity http://publius.oxfordjournals.org/ for Men and Women in the Workplace and the National Commission for Equality and Equal Opportunities between Women and Men, created in 1983 and 1984, respectively. Both agencies were limited in the pursuit of feminist policies, autonomy, and resources under rightist governments. The 1990s leftist governance brought many changes; Prime Minister established the Ministry of Equal Opportunities in 1996 and the Department of Equal Opportunities in 1997. Not only did the government create WPAs to operate from a higher institutional positioning, but it also strengthened the existing Committee and Commission. at University of Colorado on March 13, 2012

Table 2 Descriptive representation in national and meso-level legislatures in Italy, Spain, and Poland, 1990–2000s

Country Women’s descriptive representation Decade

Mid-1990s Early-2000s Mid-2000s

Italy National Representation of Women 15.1 11.5 17.3 Meso-level Representation of Women, Average 12.9 8.8 12.8 Spain National Representation of Women 22.0 28.3 36.0a Meso-level Representation of Women, Average 20.3 36.3a 41.6a Poland National Representation of Women 13.0 20.0 20.0 Meso-level Representation of Women, Average 10.4 14.2 16.9

Note: National representation pertains to lower houses of parliament and meso-level representation signifies the average representation of women across all regions after the first election of the half-decade (i.e., 1995, 2000, and 2005). See Supplementary Appendix for percent of women representatives in each meso-level legislature. aIndicates fair threshold at 30% women legislators. 90 C. D. Ortbals et al.

During this time, the national policy agencies were characterized by stronger leadership and influence; the Minister of Labor and the Minister of Equal Opportunities jointly appointed regional equality advisors; and commissions for equal opportunities were also created in most regions, provinces, and municipal- ities.5 These gains were short lived due to subsequent alternations in government and Berlusconi’s leadership (2001–2006; 2008–). In 2004, Berlusconi closed the original National Commission and replaced it with the weaker Commission for Equal Opportunities of Men and Women, and appointed inexperienced women to lead it who are not associated with the feminist movement. Downloaded from As of the mid-2000s, all regions have WPAs, with financial autonomy, although they vary by the ‘‘political color’’ of the regions (Guadagnini 2007, 178–9; Guadagnini 2005). Most regions have at least a Regional Commission for the Equal Opportunities between Man and , as well as a Regional Advisor for Equal http://publius.oxfordjournals.org/ Opportunities; however, more Commissions are feminist in their rhetoric than in actual policy actions, which include holding workshops, typically about job training, and campaigns focused on gender violence and women’s health (table 3). Although all regions have equality plans, the plans themselves and the commissions’ actions vary, for the WPAs with the strongest feminist orientations are predominantly in northern and center regions. These regions (Emilia- Romangna, , Marche, Piedmont, Toscana, and Venice) have longer established WPAs (several dating to the 1980s), female leadership, and equality plans that discuss numerous policies (including equal employment, reconciliation at University of Colorado on March 13, 2012 of work and family, and violence against women). These WPAs also included women’s organizations within their leadership structures or collaborated with them on projects, for example, through consultative activities and co-sponsored workshops and campaigns. Southern regions are moderate to low in feminist orientation and are less developed and gender-focused in rhetoric.6 In these regions (Basilicata, Calabria, Molise, Puglia, Sardegna, and Sicily), documents generically addressed ‘‘equal opportunities’’ without particular focus on women/gender and WPAs involved themselves in few policy projects and collaboration with women’s organizations. The Italian women’s movement of the late 1970s and early 1980s—with strong ties to the left- was one of Europe’s most active (Kaplan 1992), particularly so between 1974 and 1979 when it pushed for abortion reform. The movement has since fragmented into smaller groups with varying missions, addressing fertility rights, sexual harassment, and domestic violence in the 1990s (Beccalli 1996). Because there is no comprehensive list of Italian women’s organizations, the movement is often characterized as declining; however, the number of local organizations has increased (Guadagnini 2007). Earlier studies show that Italian feminism is strong in ‘‘large cities of the North and Center,’’ cities leftist in orientation, but weak in southern regions (Beccalli 1996, 159; Hellman 1987). Impact of Meso-level Institutions on Women in Politics 91

Table 3 Meso-level women’s policy agencies in Italy, Spain, and Poland

Country Number Number of Feminist policy of regions regions with with WPA equality Number of regions Number of Number of regions plan(s) expressing rhetorical regions enacting collaborating with inclusiveness for feminist policy feminist groups to feminist policy to to various various degrees various degrees degrees Downloaded from Italy, n ¼ 20 20 20 High 18 High 12 High 10 Moderate 1 Moderate 5 Moderate 6 Low 1 Low 3 Low 4

None 0 None 0 None 0 http://publius.oxfordjournals.org/ Spain, n ¼ 17 17 17 High 17 High 17 High 12 Moderate 0 Moderate 0 Moderate 5 Low 0 Low 0 Low 0 None 0 None 0 None 0 Poland, n ¼ 16 16 1 High 8 High 6 High 8 Moderate 4 Moderate 2 Moderate 1 Low 4 Low 4 Low 7 at University of Colorado on March 13, 2012 None 0 None 4 None 0 Total, n ¼ 53 53 38 High 43 High 35 High 30 Moderate 5 Moderate 7 Moderate 12 Low 5 Low 7 Low 11 None 0 None 4 None 0

Note: Feminist policy denotes three areas, Anti-Discrimination in Employment, Work-Family Balance, and Gender Violence. For regional specific results, contact the authors.

More recent data collected about organizations that run anti-violence centers, however, show change. First, the number of organizations has increased. In the early 1990s, organizations addressing violence were only found in capital cities; as of 2008, there are approximately 119 anti-violence centers located throughout the country (Forum dei Centri Antiviolenza e delle Case delle Donne 2008) and in every region except Molise (one of Italy’s smallest regions). The regions with the most organizations are still large, center, and northern regions (Emilia–Romagna: 16; Lazio: 10; Lombardy: 15; Tuscany: 11; and Venice: 11), yet the larger southern regions host multiple centers (Campania: 7; Puglia: 6; and Sicily: 8). 92 C. D. Ortbals et al.

In Italy, meso-level triangles of empowerment have not emerged consistently. The national government has not contributed to their development due to the fragility of leftist coalitions and the anti-feminism of . Evidence of emerging triangles comes from the North or Center, for more feminist-oriented WPAs tend to be in regions with higher numbers of women legislators, historically established women’s organizations, and antiviolence organizations. Italy displays few women in legislatures, notable rhetorical support for feminist policies but a lack of concrete actions, and women’s organizations spread across its regions. Downloaded from

Spain Women’s rights in Spain have advanced considerably since the country’s democratic transition (1975–1982). Spanish women gained voting rights in 1931; however, the right-wing, centralized dictatorship of Francisco Franco (1939–1975) http://publius.oxfordjournals.org/ disallowed voting for women (and men), encouraged a traditional view of women as wives and , and barred divorce and abortion. The Spanish Socialist Workers’ Party (PSOE), governing in the early 1980s, legalized abortion in some cases and pushed for gender policy; and the women’s movement, which had fought for such goals, declined thereafter.7 Decentralization empowered regional author- ities to further gender policies through meso-level women’s policy agencies, which since have spurred the creation of more women’s organizations. Spain has a long history of peripheral nationalisms seeking regional autonomy, and though the at University of Colorado on March 13, 2012 Franco regime curbed the expression of nationalisms, the Constitution of 1978 established regional autonomy and set terms for the decentralization of policies such as health and education. In the early 1980s, regions signed autonomy statutes and held their first elections. Fast-track regions gained policy powers in the 1980s, whereas slow-track regions received them in the 1990s. Nonetheless, all regions had the ability to pursue ‘‘women’s promotion’’ since the time of their autonomy. The average percentage of Spanish women in regional legislatures reached 30 percent in the early 2000s, twenty years after decentralization. Therefore, as in Italy, decentralization did not automatically prompt significant descriptive representation for women. Table 2 demonstrates that women’s representation was similarly low at the national and regional levels in the 1980s and early 1990s. Moreover, in the wake of decentralization, in 1982, there were 4.6 percent women in the national legislature and 5.7 percent women in regional legislatures on average. The dramatic rise in women in regional legislatures in the 1990s is attributable to voluntary party quotas adopted by leftist parties. PSOE socialists adopted a 25 percent quota in 1988 and a 40 percent one in 1997, whereas the center-right People’s Party (PP) has consistently rejected quotas (Valiente 2005) though informally seeking more women candidates for political office in the 1990s. Impact of Meso-level Institutions on Women in Politics 93

The 2007 Equality Law passed under PSOE mandates 40 percent women on all party election lists. The quota rules of national parties and the Equality Law apply to regional elections; thus women similarly are represented in national and regional legislatures, for the 2000 national election yielded 28.3 percent women in the Congress of Deputies and the average percent of women in regional legislatures in the early 2000s was 36.3 percent. Although national laws and party practices strongly influence regional descriptive representation, laws mandating quotas were first present in Spain in regions and women’s regional representation in the mid-2000s (41.6 percent on Downloaded from average) outpaced representation in Congress (36.0 percent). Castile-La Mancha, Balearic Islands, and the Basque Country passed regional parity quotas (50/50, men/women) in the early 2000s. Although the Constitutional Court under conservative governance in 2002 ruled regional parity laws unconstitutional http://publius.oxfordjournals.org/ (Bustelo and Ortbals 2007), it reinstated the Basque law in 2009 and Basque women are the most highly represented as of the mid-2000s (52.0 percent). Nevertheless, women in the Basque parliament do not equally occupy positions of power in the legislature, like president or commission heads, which indicates, ‘‘the presence of women ...does not necessarily imply the representation of their interests ...[and] equality policies’’ (Martı´nez and Elizondo 2006). Women’s policy agencies in Spain impacted women representation in the 1980s, a decade before women were descriptively represented in legislatures. WPAs first appeared with the 1983 establishment of the national Women’s Institute (WI) at University of Colorado on March 13, 2012 under the PSOE. The WI has the main goal of advancing gender-equality policy; is an autonomous department within the Equality Ministry as of 2008; and has been led by socialists with the exception of 1996 to 2004 (see Valiente 1995; Threlfall et al. 2005). Unlike in Italy, the national administration—through the WI-fostered development of regional WPAs by providing the model of a stable WPA with multiple policy goals. Regional WPAs were established in the late 1980s and early 1990s, with the first legally established ones in the Basque Country and Andalusia in 1988. Although WPAs did not emerge immediately upon decentralization, all regions had them within fifteen years of regional autonomy (Bustelo 2004). This is notable given that each region, with autonomy and the responsibility of ‘‘women’s promotion,’’ had to create their own gender bureaucracy. As of 2009, all but one region had published two or more equality plans, five regions had published at least four plans, and some had published additional plans addressing gender violence. All regions rhetorically supported feminist policy and had policy programs related to employment, work-life balance, and gender violence. Many employment and work-life policies are similar across regions as they stem from the European Union Equal Initiative that fights discrimination in the workplace.8 Policies unique to regions include the promotion of lesbianism in culture and lesbians’ health concerns (in Catalonia, see Platero 2008) and services 94 C. D. Ortbals et al. for women caring for elders, children, or the disabled (in Castile-La Mancha).9 All but five WPAs display high levels of collaboration with feminists. WPAs utilize leadership forums to gain organizations’ feedback and collaborate on projects, ranging from violence awareness workshops and entrepreneurship. Although regional autonomy allows conservative regions to lack a feminist ethos, resulting in feminist disapproval of WPAs and no triangle partnerships (for this argument about Galicia before 2005, see Ortbals 2007), we conclude that regional WPAs mainly demonstrate a high degree of commitment to feminist policies and organizations. Stable leftist governance, through the positive influence of the Downloaded from national Women’s Institute, set the stage for the development of regional WPAs (Threlfall et al. 2005); yet, progressive WPAs now can be found under regional governments with diverse ideologies. Catalonia, though governed since 2003 by a coalition of left-wing and left-wing nationalist parties, largely has been governed http://publius.oxfordjournals.org/ by a center-right nationalist party since its autonomy yet historically has pursued progressive policies. The women’s movement in Spain includes over 5,000 women’s organizations mobilized around diverse issues, from gender violence to rural identity (Bustelo and Ortbals 2007). The movement was weak in the country’s periphery during democratization, at which point Catalonia (10), Madrid (9), and the Basque Country (5) contained almost half of the fifty-five women’s organizations documented in feminist accounts (see Escario et al. 1996). Several regions at the time had no feminist organizations (Cantabria, Extremadura, Murcia, Navarre, and at University of Colorado on March 13, 2012 Rioja) or one (Asturias), though women in forty-seven of Spain’s fifty provinces were affiliated with housewives associations as of 1977 (Radcliff 2002).10 The large region of Andalusia and the small region of Cantabria demonstrate organizational growth in the 1980s and 1990s. In Andalusia, three feminist organizations were prominent during democratization (Escario et al. 1996), 152 women’s organizations of all kinds existed in 1989, and in 1993—five years after the establishment of the region’s WPA—600 organizations existed (and the number exceeded 1,500 in 2008). In Cantabria, as of 2007, there were ninety women’s organizations, with almost sixty of them from rural areas of the region (Cores 2007). Furthermore, 171 anti-violence centers run by women’s organizations represent all regions of Spain (Women Against Violence Europe 2008). Prominent feminist organizations and many antiviolence organizations still are located in regions with large cities (Madrid—36 and Catalonia—30); yet antiviolence organizations are dispersed into regions with few historical organizations (Cantabria—2, Extremadura—2, Murcia—4, Navarre—4, Rioja—1, Asturias—10).11 Therefore, the Spanish case shows that all triangle nodes provide women some degree of representation. Regional legislatures, however, were slow to descriptively represent women and triangles of empowerment depend more on WPAs and women’s organizations than legislatures. Meso-level ‘‘triangles’’ of empowerment Impact of Meso-level Institutions on Women in Politics 95 have emerged as women’s organizations and WPAs collaborate and as WPAs offer organizations subsidies and management training.

Poland Polish women have faced formidable challenges to equality, due to Poland’s transition from state to free market capitalism in 1989. In the early transition years, the number of women in elected office dropped sharply, women

were more likely to be unemployed or impoverished by reforms hastening Downloaded from economic liberalization and privatization, and women’s rights to social assistance and reproductive choice receded. After eight years of negotiations, the two major party blocs (rightist post-Solidarity parties and leftist successor party, the Alliance or SLD) found common ground on decentralization. In 1999, sixteen regions were created with competencies in education, health and http://publius.oxfordjournals.org/ administration, and elections were held for new regional assemblies. While Polish women received voting rights in 1918, the Catholic Church has enforced the centrality of the Polish to the nation’s survival. Under communism (1945–1989) women retained their roles in the household, but the state changed women’s roles in the public sphere by encouraging record numbers of women into the paid workforce, providing state-funded daycare, constitutionally guaranteeing women’s equality with men, and implementing 30 percent quotas for women in the then feeble Polish parliament. The state controlled the women’s at University of Colorado on March 13, 2012 movement by sanctioning only one weak organization, the Women’s League. In 1986, General Jaruzelski created a national WPA, more to recruit women to fill Sejm seats than to provide a channel for women’s participation (Robinson 1995). As of 2011, Polish women do not constitute 30 percent of legislators in the national or regional legislatures. As table 2 shows, in the mid-2000s, women comprised 20 percent of the national Sejm, and averaged only 16.9 percent of regional parliamentarians. In three rounds of regional elections (1998, 2002, and 2006), the average percentage of women in regional legislatures has remained at or below national levels, at 10.4 percent, 14.2 percent, and 16.9 percent, respectively. Prior to 2011, three Polish parties had gender quotas of at least 30 percent women candidates on lists, two of them leftist parties, Democratic Left Alliance and Labor Union, and the small centre-right Democratic Party.12 Women have done comparatively well (28.2 percent) in Malopolskie, an outlier region where elected women hail mostly from the rightist Truth and Justice Party. But women’s numbers remain low, around 10–15 percent, in Podkarpackie, Lubuskie and S´wie˜tokrzyskie (see Supplementary Appendix). Polish women also face masculine gendered regional legislatures in which male legislators view female legislators, and female legislators view themselves, as political outsiders (Rincker 2009). 96 C. D. Ortbals et al.

National and regional women’s policy agencies emerged in the last decade, but their lack of institutionalization undercuts potential for women’s representation. The communist-era WPA, or Women’s League, was not feminist neither was the Office of Women’s Affairs operating in the early 1990s (Robinson 1995). The first post-communist, feminist WPA emerged in 2001 under leftist SLD Premier Leszek Miller (2001–2004) who created a cabinet-level Plenipotentiary for the Equal Status of Women and Men. In 2003, SLD and Labour Union (UP) legislators and women’s organizations created 16 regional WPAs. The regional WPA leaders were to be nominated by regional governors by July 1, 2004. As table 3 shows, after one Downloaded from year of operation, only one regional WPA in Poland had developed an equality plan. In eight of sixteen regions, however, the regional WPAs pursued feminist policies. These regions implemented a national WPA initiative ‘‘To Overcome Violence,’’ and some of them also created unique programs such as antipoverty http://publius.oxfordjournals.org/ policies tailored to Roma women (Opolskie) and a Forum for Women (S´la˜skie), at which a female legislator noted ‘‘we discussed violence in the family, women’s unemployment ...we could talk openly about it, which we cannot do in the regional parliament’’ (Rincker 2009). The regions with feminist policy proposals tended to be left or center-left, but more importantly, had strong feminists as WPA directors who advanced programs to change gender norms. Rightist regions were slower to nominate their WPA leaders. The shift in national governance from the leftist SLD to rightist Truth and Justice (in 2005) and Civic Platform (in 2008) dismantled many regional WPAs in at University of Colorado on March 13, 2012 Poland. The structure and mission of all regional WPAs shifted after 2005 when the Truth and Justice Prime Minister Marcinkiewicz abolished the cabinet-level WPA, and demoted it under the Ministry of Labor and Social Policy. Even after the 2008 election of center-right Civic Platform government, some regions have been slow to nominate a WPA director yet other regions have continued their efforts. As of 2010, Poland has a Vice-Minister for Women, Family, and Counteracting Discrimination, underneath the Minister of Labor and Social Policy. Regional initiatives, such as a plan on Violence Against Women, are promoted through Provincial offices of Social Policy. Poland thus demonstrates that in a weakly decentralized system, rightist Prime Ministers can dismantle or thwart the progress of innovative meso-level WPAs. Polish women’s organizations are growing and dispersing, despite the legacy of state socialism and restrictions on civil society throughout the period of 1945–1989. Many Polish women’s organizations temporarily coalesced in the early 1990s to protest new laws outlawing abortion for the first time in forty-five years. These same organizations suffered attrition in the mid-1990s as their members struggled to face the burdens of transition to a competitive market economy. But by the 2000s, women’s organizations in Western and Central Poland focus mainly on alimony/child support payments, economic rights, work-family issues, and abortion Impact of Meso-level Institutions on Women in Politics 97 rights. In eastern Poland, organizations are primarily Women’s Circles, or rural women’s social groups that promote traditional gender roles. Still, more women’s organizations are present after decentralization (105 groups, 2001) than before (80 groups, 1997), and they are more dispersed outside of the capital Warsaw. In 1997, prior to decentralization, 60 percent of women’s groups were located outside of Warsaw. By 2001, 64 percent were located outside of Warsaw. The first antiviolence center in Poland was opened in the mid 1990s, and by 2008 there were 33 organizational antiviolence centers with at least one in each region. Thus, decentralization is associated with organizational growth outside the capital city. Downloaded from In fact, the Amazons and Europa Donna organizations have formed administrative branches in regions outside the capital. The triangle of empowerment in Polish regions is not yet institutionalized. Where strong feminist regional WPA directors were appointed, they worked http://publius.oxfordjournals.org/ effectively with existing women’s organizations and linked them to regional and national legislators on policy projects. The 2005 elections negatively affected the national WPA, and it has not rebounded under the centrist Civic Platform government since 2008. The western and more leftist regions of Poland generally tend to elect more women, and contain more women’s organizations with feminist orientation. It remains to be seen whether Poland’s 2004 European Union accession will pressure the Polish state to put regional WPAs on solid institutional footing. at University of Colorado on March 13, 2012 Conclusion Our analysis of meso-level representation in Spain, Italy, and Poland shows that decentralization produces potential avenues for women in politics—but some meso-level institutional settings better represent women citizens than others. In the short term, meso-level legislatures did not benefit women as anticipated. With the exception of Spain during the 2000s, meso-level legislatures did not represent women at or above the U.N. understanding of fair representation at 30 percent and women’s representation in meso-level legislatures lagged representation in national legislatures. Even as descriptive representation has increased, increased substantive representation and women’s leadership in legislatures has not transpired, as cited in the Basque region of Spain and in Polish regions. In contrast, meso-level women’s policy agencies enhance women’s representation by furthering feminist policy and engaging in feminist collaborations, even in the short term. All regions in Italy, Spain, and Poland developed WPAs that almost unanimously communicate high levels of rhetorical support for feminist policies and with the exception of most Polish regions pursue equality plans. Further, forty-two of fifty-three regions studied pursue a high (35) or moderate (7) degree of policy action regarding equal employment, reconciliation of work and family, and violence against women. Moreover, we found that women’s organizations established themselves and worked 98 C. D. Ortbals et al. outside of the capital city, moving into peripheral areas previously lacking this venue for women’s representation. We counted a total of 323 antiviolence organizations in these countries, which greatly exceeds the number of organizations in existence at the start of their decentralizations. While we did not find consistent collaboration across the triangle of empowerment, we identified examples where committed actors in two or three of the nodes of the triangle worked in concert. This is significant because Italy, Spain, and Poland have scored low in previous studies for interaction between women’s organizations and WPAs at the national level (Stetson and Mazur 1995). Downloaded from In the S´la˜skie region in Poland, we found that legislators, WPAs, and activists came together to discuss and address women’s needs. Spanish regional WPAs remarkably demonstrate moderate to high collaboration with women’s organizations, and the

number of Spanish organizations has grown substantially through WPA assistance http://publius.oxfordjournals.org/ (see Ortbals 2010). Moreover, only four of twenty Italian regions display low levels of collaboration with women’s organizations; and only seven of sixteen Polish regions rank low for collaboration. Even without full triangle partnerships, representation is evident when strong ‘‘critical actors’’ propel meso-level units toward greater representation of women. For instance, feminist leadership in Polish WPAs was critical to WPA actions and the historical feminist movement in Italy was key to emerging triangles in Italy. Likewise, advances in descriptive

representation in Spain can be attributed to national parties that voluntarily at University of Colorado on March 13, 2012 adopted quotas, while, in Italy and Poland, national parties and laws either facilitated quotas or disallowed them temporarily as in Italy. Throughout the case studies, variations in decentralization and women’s representation emerged that deserve further discussion. Spain stands out as an exemplary state in that all regions have 30þ percent women in legislatures, many women’s organizations, and WPAs that support feminist policies. We suggest that stronger decentralization and stable leftist governance at the national level over many years (1982–1996) benefited Spanish women, whereas Polish and Italian regional WPAs have been less influential due to their countries’ national insti- tutional and political instability and rightist downgrading of WPAs. For instance, in Poland, as expected in unitary states, the central government can repossess decentralized policy competencies, as it did with the function of women’s policy agencies in 2005, thereby stymieing regional feminist policymaking. On the other hand, Spain represents the strong decentralization of many federal states, thus allowing for meso-level institutional and policy stability no matter the political color of the national government. Therefore, we emphasize that future studies of women in meso-level institutions must account for the impact of national level variables, the degree and permanence of decentralization, and leftist governance. Impact of Meso-level Institutions on Women in Politics 99

We also find that the European Union has encouraged all three previously lagging, Catholic countries to improve women’s representation, whether through WPA creation or equal opportunity mandates. Increased meso-level representation for women therefore has transpired in a multilevel way. Feminist policies are the result of enmeshed policy networks as, for example, the EU’s EQUAL initiative has formed the basis of some region’s responses to equal opportunities. Additionally, regions are innovators whose practices ‘‘trickle up’’ to higher levels. In some Spanish and Italian regions, regional quota laws increased descriptive representation in legislatures. Spain since has passed a national quota and conservative regions will Downloaded from benefit from it. What is more, the EU’s gender initiatives emphasize women’s political representation. This demonstrates that in the European context, no matter the particular type of decentralization, or federation, for that matter, supranational laws and practices will impact women in the meso-level sphere. http://publius.oxfordjournals.org/ Additional variables that help determine women’s representation close to home are variables at the meso-level itself. Our study and secondary sources indicate that regional political party governance, level of centralization of political parties, socio-economic development of region, local critical actors, local political culture, and preexisting mobilization of women’s organizations affect women’s meso-level representation. Findings herein affirm the positive influence of leftist governance, though illustrative examples, such as Catalonia, an economically advanced region in Spain, suggest that leftist governance is not a necessary condition for progressive gender policies. Moreover, Chaney, MacKay, and McAllister note that the at University of Colorado on March 13, 2012 mobilization of women’s groups in the United Kingdom prior to devolution ‘‘ensure[d] gender equality was at the heart of the emerging devolution proposals’’ (2007, 36). These examples suggest that future research must include regional-level economic, political, and women’s movement variables as indicators of women’s representation and collaboration in the triangle of empowerment. In summary, our research shows that the establishment of meso-level government in Spain, Italy, and Poland created new institutions in which women participate. In that sense, decentralization, particularly in the European context, increases the overall representation of women in politics compared to that of centralized polities. Meso-level institutions, however, do not automatically beget high levels of women’s representation. From this main finding, we derive several lessons for federalism and decentralization scholars and policymakers. First, for federalism scholars, we demonstrated comparative evidence that meso-level institutions in unitary states, as in federal states, present advantages and disadvantages, or, in other words, a ‘‘mixed picture’’ (Gray 2002, 38). Politics close to home spawns women’s movements and regional gender bureaucracies in decentralized, unitary states, but their effectiveness partially depends upon the stability of decentralized institutions, which cannot be assumed in many cases of decentralization outside of federations. The second lesson pertains to variation in 100 C. D. Ortbals et al. the meso-level sphere. We show that most similar decentralized cases vary quite a bit in representing women, thus reflecting findings by federalism scholars who show that women’s representation in federations vary as well (Sawyer 2002; Chappell 2002). If outcomes in federal states vary amongst themselves as do decentralized states’ outcomes, scholars must conduct more comparative studies to tease out whether meso-level institutional design is driving causal outcomes or whether other variables are equally or more salient (Gray 2006, 38). The third lesson, for policymakers, is that decentralization must be accompanied by gender quotas and meso-level WPAs in order to advance women’s representation in the Downloaded from early years of decentralization, for women’s organizations have greater resources and incentives to form in peripheral regions if they can appeal to meso-level institutions like WPAs. Thus, we conclude that citizens from marginalized groups, though presented with opportunities through decentralization, will most likely http://publius.oxfordjournals.org/ benefit from sub-state institutions when they are specifically designed to be representative.

Supplementary Data Supplementary data can be found at www.publius.oxfordjournals.org.

Notes at University of Colorado on March 13, 2012 1. Peripheral nationalisms in Spain still motivate debates about greater regional autonomy. 2. These countries are not Europe’s wealthiest, yet Spain and Italy have similar GDP per capital (table 1). A same-time comparison between Catholic, middle-tier economies is not possible. 3. We acknowledge electoral rule changes over time. Italy had a mixed electoral system from 1993 to 2005 with 75 percent of seats allocated using SMD and 25 percent PR, but as of 2005 it returned to closed-list proportional representation system, now with a 20 percent threshold for coalitions, 3 percent threshold for any party within a coalition, and 8 percent threshold for a single party. District magnitude in Spain ranges from 3 to 34 with a 3 percent electoral threshold. Poland has a 5 percent electoral threshold for parties and 8 percent threshold for coalitions. As per gender quotas, Italy used mandated quotas in the Chamber of Deputies during the 1994 election and the 1995 regional elections, some regions mandated quotas in the 2005 regional elections, and some parties opted for voluntary quotas in the 2006 national election. Spain passed the Equality Law in 2007, after which parties were required to construct electoral lists (for European, national, regional, and municipal elections) with a minimum of 40 percent (or maximum of 60 percent) women. Three Polish parties have 30 percent voluntary gender quotas. Three Polish parties [had] voluntary gender quotas. As of January 2011, Polish President Komorowski signed into law a quota requiring a minimum of 35 percent of either gender on electoral lists for gmina, powiat, Sejm, and European Parliament elections. 4. Other organizations exist, yet anti-violence organizations are cross-nationally reported. Impact of Meso-level Institutions on Women in Politics 101

5. In 1991, equality advisors were created but were poorly delineated and funded. Reforms in 1999 and 2000 gave advisors guidelines for monitoring the workplace and E10 million annually. 6. However, Abruzzo and Campania address feminist issues and collaborate with organizations. 7. As of 2010, abortion is legal until the fourteenth week of pregnancy without restrictions. Previously, abortion was legal in cases of rape, fetus malformations, and when a mother’s physical or psychological health is seriously imperilled. Divorce was legalized in 1981. Downloaded from 8. The European Social Fund funds this initiative. 9. Spanish regions led the way to domestic partnerships in the late 1990s, before the national administration passed a 2005 law (Ley 13/2005) permitting same-sex marriages (Platero 2008). 10. Provinces exist between the regional and municipal levels. Housewives associations advocated for family and consumer issues. http://publius.oxfordjournals.org/ 11. Prominent groups include Federacio´n de Organizaciones Feministas del Estado Espan˜ol, Federaci0n de Mujeres Progresistas, Asociacio´n de Mujeres Juristas, and Coordinadora Espan˜ol del Lobby Europeo de Mujeres. Asociacio´n de Mujeres Juristas is located in Madrid, but the Federaci0n de Mujeres Progresistas is organized by branches in each region. 12. President Komorowski signed a law on January 5, 2011 requiring that ‘‘the number of candidates of either gender on the electoral lists may not be lower than 35 percent of the overall number of candidates on this list ...and for lists that include 3 candidates, there at University of Colorado on March 13, 2012 must be at least one candidate of each gender.’’ (Global Database of Gender Quotas www.quotaproject.org, Accessed June 30, 2011).

References Alvarez, Sonia. 1990. Engendering Democracy in Brazil. Princeton: Princeton University Press. Banaszak, Lee Ann, Karen Beckwith, and Dieter Rucht. 2003. Women’s movements facing the reconfigured state. Cambridge and New York: Cambridge University Press. Beccalli, Bianca. 1996. The modern women’s movement in Italy. In Mapping the women’s movement: Feminist politics and social transformation in the north, ed. Monica Threlfall. 152–183. London: Verso Press. Beckwith, Karen. 2000. Beyond compare? Women’s movements in comparative perspective. European Journal of Political Research 37 (4): 431–468. Bonomi, Genny, Giorgio Brosio e Maria, and Laura Di Tommaso. 2006. How Italian electors react to gender quotas? A random utility model of voting behaviour. Working Paper Series, No. 09/2006, Universita` di Torino, at http://ideas.repec.org/p/uto/dipeco/200609. html (accessed November 18, 2008). Brady, Henry, Kay Lehman Schlozman, and Sidney Verba. 1999. Prospecting for participants: Rational expectations and the recruitment of political activists. American Political Science Review 93 (1): 153–168. 102 C. D. Ortbals et al.

Breton, Albert. 2000. Federalism and decentralization: Ownership rights and the superiority of federalism. Publius: The Journal of Federalism 30 (2): 1–16. Bustelo Ruesta, Marı´a. 2004. La evaluacio´n de las polı´ticas de igualdad de ge´nero en Espan˜a. Madrid: La Catarata. Bustelo, Marı´a, and Candice D. Ortbals. 2007. The evolution of Spanish . A fragmented landscape. In Changing state feminism: Women’s policy agencies confront shifting institutional terrain, ed. Joyce Outshoorn, and Johanna Kantola. 201–223. New York: Palgrave Macmillan.

Cain, Bruce E., Russell J. Dalton, and Susan E. Scarrow, eds. 2003. Democracy transformed? Downloaded from Expanding political opportunities in advanced industrial democracies. Oxford: Oxford University Press. Celis, Karen, and Alison Woodward. 2003. Flanders: Do it yourself and do it better? Regional parliaments as sites for democratic renewal and gendered representation. In Regional http://publius.oxfordjournals.org/ institutions and governance in the European Union, ed. Jose´ Magone. 173–191. Westport, CT: Praeger. Chaney, Paul, Fiona MacKay, and Laura McAllister. 2007. Women, politics and constitutional change: The first years of the national assembly for Wales. Cardiff, Wales: University of Wales Press. Chappell, Louise A. 2002. Gendering government: Feminist engagement with the state in Australia and Canada. Vancouver: UBC Press. Childs, Sarah, and Mona Lena Krook. 2006. Should feminists give up on critical mass? A contingent yes. Politics & Gender 2 (4): 522–530. at University of Colorado on March 13, 2012 Chin, Mikyung. 2004. Reflections on women’s empowerment through local representation in South Korea. Asian Survey 44 (2): 295–315. Cores, Mariana. 2007. Mujeres asociadas, la satisfaccio´n de compartir. Diario Montan˜es 18 March 2007. Darcy, R., Susan Welch, and Janet Clark. 1994. Women, elections and representation. Lincoln, NE: University of Nebraska Press. Delegacio´n de la Mujer. 2003. Compartiendo espacio con las mujeres de Sevilla, Memoria de actuaciones 1999–2003. Spain: Seville. Elazar, Daniel. 1991. Exploring federalism. Tuscaloosa, AL: University of Alabama Press. Escario, Pilar, Ine´s Alberdi, and Ana Ine´sLo´pez-Accotto. 1996. Lo personal es polı´tico: El Movimiento Feminista en la Transicio´n. Madrid: Ministerio de Asuntos Sociales. Falleti, Tulia G. 2005. A sequential theory of decentralization: Latin American cases in comparative perspective. American Political Science Review 99 (3): 327–346. Ferree, Myra Marx, and Aili Mari Tripp. 2006. : Transnational women’s activism, organizing, and human rights. New York: New York University Press. Fodor, Eva, Christy Glass, Janette Kawachi, and Livia Popescu. 2002. Family policies and gender in Hungary, Poland and Romania. Communist and Post-Communist Studies 35 (4): 475–490. Impact of Meso-level Institutions on Women in Politics 103

Folbre, Nancy. 2001. The invisible heart: Economics and family values. New York: The New Press. Forum dei Centri Antiviolenza e delle Case delle Donne, at http://www.women.it/casadonne/ comecitrovi/ (accessed November 22, 2008). Gelb, Joyce, and Marian Lief Palley. 1982. Women and public policies. Princeton, NJ: Princeton University Press. Gray, Gwen. 2006. Women, federalism and women friendly policy. Australian Journal of Public Administration 65 (1): 25–45. Downloaded from Quota Project: Global Database of Quotas for Women. www.quotaproject.org. [accessed May 12, 2011]. Guadagnini, Marila. 2005. State feminism in Italy from the ‘’ to Berlusconi: Few achievements in a changing political context. Paper prepared for the ECPR. Granada –

14–19. http://publius.oxfordjournals.org/ ———. 2007. The reform of the state in Italy. In Gendering the state in the age of globalization, ed. Melissa Haussman, and Birgit Sauer. 169–188. Lanham: Rowman and Littlefield Publishers. Guadagnini, Marila, and Alessia Dona`. 2007. Women’s policy machinery in Italy between European pressure and domestic constraints. In Changing state feminism, ed. Joyce Outshoorn, and Johanna Kantola. New York: Palgrave Macmillan. Halsaa, Beatrice. 1998. A strategic partnership for women’s policies in Norway. In Women’s

movements and public policy in Europe, Latin America, and the Caribbean, ed. at University of Colorado on March 13, 2012 Geertje Lycklama a` Nijeholt, Virginia Vargas, and Wieringa Saskia. 167–189. New York and London: Garland Publications. Haussman, Melissa. 2005. Abortion politics in North America. Boulder, CO: Lynne Rienner Publishers. Hellman, Judith Adler. 1987. Women’s struggle in a workers’ city: Feminist movements in Turin. In The women’s movements of the United States and Western Europe: Consciousness, political opportunity, and public policy, ed. Mary Fainsod Katzenstein, and Carol McClurg Mueller. 111–131. Philadelphia: Temple University Press. Holli, Anna Maria, and Johanna Kantola. 2007. State feminism Finnish style: Strong policies clash with implementation problems. In Changing state feminism, ed. Joyce Outshoorn, and Johanna Kantola. 82–101. New York: Palgrave Macmillan. Interparliamentary Union. 2009. Women in national parliaments, at http://www.ipu.org/ wmn-e/world.htm (accessed September 30, 2009). Kaplan, Gisela. 1992. Contemporary Western European feminism. New York: New York University Press. Krook, Mona. 2009. Quotas for women in politics: Gender and candidate selection reform worldwide. Oxford: Oxford University Press. Litvack, Jennie, Junaid Ahmad, and Richard Bird. 1998. Rethinking Decentralization in Developing Countries. Washington, DC: The World Bank. 104 C. D. Ortbals et al.

Litvack, Jennie, and Jessica Seddon. 2002. Decentralization briefing notes. World Bank Institute Working Papers. Lombardo, Emanuela, and Petra Meier. 2008. Framing gender equality policy in European Union political discourse. Social Politics: International Studies in Gender, State & Society 15 (1): 101–129. Lycklama a` Nijeholt, Virginia Vargas, and Saskia Weringa, eds. 1998. Women’s movements and public policy in Europe, Latin America, and the Caribbean. New York and London: Garland Publishing, Inc. Macaulay, Fiona. 2006. Gender politics in Brazil and Chile. Basingstoke, United Kingdom: Downloaded from Palgrave Macmillan. Martı´nez Herna´ndez, Eva, and Arantxa Elizondo Lopetegi. 2006. Participacio´n Polı´tica de las Mujeres: Presencias y Ausencias en la Polı´tica Vasca. Vasconia 35: 491–502, at www.euskomedia.org/PDFAnlt/vasconia/vas35/35491502.pdf. http://publius.oxfordjournals.org/ Marks, Gary. 1993. Structural policy and multilevel governance in the European community. In The state of the European community: The Maastricht debate and beyond, ed. Alan W. Cafruny, and Glenda G. Rosenthal. 391–411. Boulder: Lynne Reiner Publishers. Mazur, Amy G. 2002. Theorizing feminist policy. Oxford: Oxford University Press. Mazur, Any G., and Dorothy E. McBride. 2008. State feminism. In Politics, gender, and concepts: Theory and methodology, ed. Gary Goertz, and Amy G. Mazur. 244–269. New York: Cambridge University Press. at University of Colorado on March 13, 2012 Montoya, Celeste. 2009. International initiative and domestic reforms: European Union efforts to combat violence against women. Politics & Gender 5 (3): 325–348. Moreno Mı´nguez, Almudena. 2006. The persistence of male breadwinner model in southern European countries in a compared perspective: Familism, employment and family policies. The Annals of Marie Curie Fellowship Association, at http://www.mariecurie.org/annals/ volume4/soc1.pdf (accessed March 10, 2009). Northern Ireland Assembly. 2008. ‘‘EU Sub-National Governments.’’ http://www.niassembly .gov.uk/io/research/2008/12108.pdf (accessed July 15, 2010). O’Neill, Kathleen. 2003. Decentralization as an electoral strategy. Comparative Political Studies 36 (9): 1068–1091. Ortbals, Candice D. 2007. Jumbled women’s activism: Meso-level and international influences on Galician equality politics. International Feminist Journal of Politics 9 (3): 359–378. ——— 2008. Meso-level politics in Spain: New avenues for women’s activism and policymaking. Politics & Gender 4 (1): 93–119. ———. 2009. Feminized decentralization: Evaluating meso-level women’s policy agencies and women’s representation in Spain. In Democracy and institutional development: Spain in comparative theoretical perspective, ed. Bonnie N. Field, and Kerstin Hamann. Palgrave MacMillan. Impact of Meso-level Institutions on Women in Politics 105

——— 2010. The potential of local women’s associations in Andalusia: Pursuing culture, enriching lives, and constructing equality. South European Society and Politics 15 (2): 203–223. Ostrom, Elinor. 1996. Crossing the great divide: Coproduction, synergy, and development. World Development 24 (6): 1073–1087. Outshoorn, Joyce, and Johanna Kantola. 2007. Changing state feminism. New York: Palgrave Macmillan. Parry, Janine A. 2005. Women’s policy agencies, the women’s movement, and political representation in the USA. In State feminism and political representation, ed. Downloaded from Joni Lovenduski. 239–259. Cambridge, MA: Cambridge University Press. Platero, Raquel. 2008. Queremos las lesbianas ser mujeres? Las lesbianas a los ojos del feminismo de Estado: representaciones y retos de las sexualidades no normativas. In Lesbianas, Discursos, y Representaciones, ed. Raquel Platero. 173–190. Barcelona: http://publius.oxfordjournals.org/ Melusina. Prud’homme, Remy. 1995. The dangers of decentralization. World Bank Research Observer 10 (2): 201–220. Putnam, Robert. 1993. Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press. Radcliff, Pamela. 2002. Citizens and housewives: The problem of female citizenship in Spain’s transition to democracy. Journal of Social History 36 (1): 77–100. Rai, Shirin. 2003. Mainstreaming gender, democratizing the state? Institutional mechanisms. at University of Colorado on March 13, 2012 New York: Transaction Publishers. Rincker, Meg E. 2009. Masculinized or marginalized: Decentralization and women’s status in regional Polish institutions. Journal of Women, Politics & Policy 30 (1): 46–69. Rincker, Meg E., and Candice D. Ortbals. 2009. Leaders or laggards: Engendering sub-national governance through women’s policy machineries in Spain and Poland. Democratization 16 (2): 269–297. Robinson, Jean. 1995. The Liga Kobiet in Poland. In Comparative state feminism, ed. Dorothy McBride Stetson, and Amy G. Mazur. 203–220. Thousand Oaks, CA: Sage Publications. Rodden, Jonathan. 2004. Comparative federalism and decentralization: On meaning and measurement. Comparative Politics 36 (4): 481–500. Rom, Mark C., Paul E. Peterson, and Kenneth F. Scheve Jr. 1998. Interstate competition and welfare policy. Publius: The Journal of Federalism 28 (3): 17–37. Rowbotham, Sheila. 1996. Introduction: Mapping the women’s movement. In Mapping the women’s movement: Feminist politics and social transformation in the north, ed. Monica Threlfall. 1–16. New York: Verso. Rule, Wilma, and Joseph F. Zimmerman. 1994. Electoral systems in comparative perspective: Their impact on women and minorities. Westport, CT: Greenwood Publishing Group. 106 C. D. Ortbals et al.

Sauer, Birgit. 2003. Revisiting state feminism in the context of globalization. Paper presented at Gender, Globalization, and State Transformation: New Concepts and Approaches in Political Research Workshop, Universita¨t Bremen/Hanse Wissenschaftskolleg, September 27–30, 2003. Sawyer, Marian. 2002. Waltzing Matilda: Gender and Australian political institutions. In Australia Reshaped: 200 Years of Institutional Transformation, ed. Geoffrey Brennan, and Francis G. Castles. 148–180. Cambridge, MA: Cambridge University Press. Siahaan, Asima Yanty. 2003. The politics of gender and decentralization in Indonesia.

Budapest, Hungary: Open Society Institute. Downloaded from Stetson, Dorothy, and Amy Mazur. 1995. Comparative state feminism. Thousand Oaks, CA: Sage Publications. Swianiewicz, Pawel. 2002. Reforming local Government in Poland: Top-down and bottom up processes. Paper presented at the Joint Conference of IPSA and the German Political http://publius.oxfordjournals.org/ Science Association Workgroup on Local Government Studies, Stuttgart, September 26–27, 2002. Teghtsoonian, Katherine, and Louise Chappell. 2008. The rise and decline of women’s policy machinery in British Columbia and New South Wales: A cautionary tale. International Review of Political Science 29 (1): 29–51. Threllfall, Monica, Christine Cousins, and Celia Valiente Fernandez. 2005. Gendering Spanish democracy. London: Routledge Press. Valiente, Celia. 1995. The power of persuasion: The Instituto de la Mujer in Spain. In at University of Colorado on March 13, 2012 Comparative state feminism, ed. Dorothy McBride Stetson, and Amy G. Mazur. 221–236. Thousand Oaks, CA: Sage Publications. ———. 2005. The women’s movement, gender equality agencies and central-state debates on political representation in Spain. In State feminism and political representation, ed. Joni Lovenduski. 174–194. Cambridge, MA: Cambridge University Press. Vengroff, Richard, Zsolt Nyiri, and Melissa Fugiero. 2003. Electoral system and gender representation in sub-national legislatures: Is there a national sub-national gender gap? Political Research Quarterly 56 (2): 163–173. Vickers, Jill. 2010. A two-way street: Federalism and women’s politics in Canada and the United States. Publius: The Journal of Federalism 40: 412–435. Watts, Ronald L. 1999. Comparing federal systems. Montreal: McGill-Queen’s University Press. Waylen, Georgina. 1994. Women and democratization: Conceptualizing gender relations in transition politics. World Politics 46 (3): 327–354. ———. 2007. Engendering transitions: Women’s mobilization, institutions and gender outcomes. Oxford: Oxford University Press. Impact of Meso-level Institutions on Women in Politics 107

Weldon, S. Laurel. 2002. Protest, policy, and the problem of violence against women: A cross-national comparison. Pittsburgh: University of Pittsburgh Press. Wolfe, Alan. 2008. The coming religious peace. The Atlantic. http://www.theatlantic.com/ doc/200803/secularism (accessed June 20, 2011). Women against Violence Europe. 2009. http://www.wave-network.org/ (accessed May 28, 2009). Downloaded from http://publius.oxfordjournals.org/ at University of Colorado on March 13, 2012